Amazon’s significant investment in the U.S. artificial intelligence company Anthropic is under formal investigation by the U.K.’s Competition and Markets Authority (CMA). The CMA announced on Thursday that it has initiated a “Phase 1” investigation to determine if Amazon’s partnership with Anthropic could potentially harm competition within the U.K. market. This decision follows an initial review that provided the CMA with enough information to proceed with a formal inquiry. The regulator now has up to 40 working days to decide if a more detailed “Phase 2” investigation is necessary.
Amazon’s investment in Anthropic, finalized in March, totaled $4 billion. This included an initial $1.25 billion equity stake in September, followed by an additional $2.75 billion. As part of the agreement, Anthropic’s advanced language models will be available on Amazon’s Bedrock platform for developing generative AI applications. These models will also be trained and deployed using Amazon’s custom AI chips, developed by its Amazon Web Services division.
In response to the CMA’s investigation, an Amazon spokesperson expressed disappointment, stating that the partnership with Anthropic does not pose any competition issues or meet the CMA’s criteria for review. The spokesperson emphasized that Amazon’s investment aims to enhance choice and competition in AI technology and noted that Amazon does not hold any decision-making power or board seats at Anthropic. They also highlighted that Anthropic is free to collaborate with other providers and has multiple partners.
An Anthropic representative reiterated the company’s independence, stating that their strategic partnerships and investor relationships do not compromise their corporate governance or freedom to partner with others. They welcomed the opportunity to cooperate with the CMA to clarify the nature of Amazon’s investment and their commercial collaboration.
The Amazon-Anthropic deal is not the only one under scrutiny by U.K. regulators. The CMA is also examining Microsoft’s substantial investment in AI leader OpenAI, although it has not yet decided whether to launch a Phase 1 investigation into that partnership. In the U.S., the Federal Trade Commission has requested information from Microsoft, Amazon, Google, OpenAI, and Anthropic regarding their recent investments and partnerships.
Some smaller tech companies have criticized large tech firms for acquiring stakes in key AI startups to gain closer access to advanced AI systems. Matt Calkins, CEO of enterprise software company Appian, argued that merely acquiring data and stakes in AI startups does not guarantee success. He suggested that the AI market is diverse, with different algorithms serving various purposes, and their value depends significantly on how data is integrated into them.